epithelial and morphological adaptations to dry habitats: a preliminary survey of adaptive trait variation among Colombian dry forest anurans.

Thesis dissertation presented by:

Juan Salvador Mendoza Roldán

Director:

Dr. Andrew J. Crawford.

Universidad de Los Andes, Bogotá, Colombia.

2014.

Resumen: Los anuros poseen una organización dermal simple que ha evolucionado para solucionar los problemas atribuidos a la terrestrealizacion. La innovación estructural como la aparición de glándulas con un amplio espectro de secreciones y la presencia de regiones especializadas, altamente vascularizadas han permitido la supervivencia de los anuros adultos en ambientes secos, dominados por altas temperaturas y la presencia de sustratos y corrientes de aire desecantes. Estas especies muestran adaptaciones tegumentarias para la perdida de agua, que van desde la presencia de osteodermos y co-osificación craneal hasta el uso de secreciones de origen lipídico. Estas adaptaciones morfológicas se encuentran acopladas con rasgos etológicos y ecológicos que configuran la estrategia adaptativa de la especie. La presente contribución se enfoca en la caracterización básica de las estructuras del tegumento, por medio de microscopia de luz. Se comparó la variación de caracteres discretos entre poblaciones y en algunos casos especies hermanas presentes en hábitats húmedos y secos. Se probó el efecto de algunas variables climáticas sobre el tamaño corporal para establecer el valor adaptativo de las diferencias intra e inter especificas existentes entre proporciones de la tibia y el cráneo, medidas relacionadas con la relación superficie y volumen. Las comparaciones realizadas entre poblaciones hermanas de distintos orígenes geográficos y de hábitat se realizaron para describir la relación existente entre algunos aspectos de la morfología externa, histología características pluviométricas, haciendo énfasis en la biota anfibia de uno de los ecosistemas terrestres más amenazados de Colombia, el bosque seco tropical. Abstract: Anurans possess a very simplified dermal organization, which has evolved to solve the basic problems of terrestrialization. Structural innovation, presence of specialized highly vascularized regions and gland sets with a wide diversity of secretions, have allowed adult anurans to survive in desiccating environments that are dominated by dry substrates and air currents, in places with elevated day temperatures; geographically this places may be generalized by having extended periods with no rainfall thus dominating dry conditions. These species show interesting integumentary adaptations to avoid water loss that range from the presence of osteoderms and skin co ossification to the use of lipid based impermeable secretions, generally these morphological adaptations are coupled with behavioral traits that together configure the adaptive strategy of the species. The present contribution focuses on the examination of anatomical components configuring the dermal organization of some Caribbean dry forest species, and by means of light microscopy characterize the basic structure of species integument to compare variation of discrete traits among conspecific populations and in some cases pairs of dry and wet habitat sister species. Body size variation was tested to establish the adaptive value in water economy conferred by body proportions related to the total surface to volume ratio. Body proportions included the analysis of variation between sibling populations, where total lengths (SVL) were contrasted with Tibial length and Craneal width. Comparisons among conspecific populations from different geographical and habitat related origin were made in order to describe the basic relation between external morphology, histology and the habitat rainfall category (Dry or Wet), focusing on the biota from one of Colombia´s most threatened land ecosystem the Seasonally dry tropical forest.

Introduction:

The colonization of terrestrial habitats by begun in the end of the Devonian period 360 million years ago, when freshwater Rhipidista with lobed fins, migrated from pond to pond during the dry season, behavior that aided in the survival of water dependent , with physiological boundaries for free dwelling on terrestrial ecosystems (Toledo et al. 1993; Romer, 1959). Dehydration always will be the earliest of amphibian problems; many fossil species show the presence of scales and bony plates that favored water retention (Colbert, 1969 in Toledo et al. 1993). Present amphibians are poorly adapted to strict terrestrial life, their skin is a very simple dermal integument that does not generally serve as a barrier to the flow of water from and towards the amphibian body, creating two mayor selective pressures crucial in the evolution of modern amphibians; aquatic species tend to hydrate and loose inner solutes and terrestrial forms tend to dehydrate by means of evo transpiration (Porter 1972).

The morpho-physiological interaction of amphibians with their abiotic environment is a complex and dynamic system of related process, arid habitats such as dry forests, deserts and open shrub lands and savannas, impose a rigorous environmental filter that has caused morphological, physiological and behavioral evolution of a wide diverse of adaptive traits that function in synergy to configure independent overall adaptive strategies for each member of the anuran community (Toledo et al. 1993; Duellman and Trueb, 1986). Amphibians don not drink the water required for metabolic function (Ex. ), this water penetrates their bodies, principally by the way of the integument , special zones for rehydration are present in different zones of the amphibian body, reason why skin permeability to water differs from one part of the to the other (Toledo et al. 1993). This author also concludes that Inter and intra specific variation among skin traits may be related to adaptation for a particular environment, and this variation may confer structural differences in the integuments. Canziani and Cannata (1980) have shown that arid region Ceratophrys ornata, has a smooth ventral skin except in the pelvic region, where it is granular, on the other hand Individuals from moist temperate climates have uniformly granular ventral skin; while dehydrating arid area may lose less water, but moist area frogs are better rehydrating by the presence of a granular skin.

Skin thickness and the number of epidermal skin layers vary across amphibian species, in the process of keratinization a process related with the aquatic or terrestrial environment. Interspecific analyses have shown that species from the African genus Ptcychadena have an inverse relationship between body size and skin width, with the largest having the thinnest skin (Le Quang Trong, 1975). Other genus such as the African Phrynobatrachus, show variable skin thickness related to diversity of habitat, forest species have thinner skins than do savanna species of the same size, and ubiquitous species have a skin thickness intermediate to these two (Le Quang Trong, 1971). Skin gland density per square millimeter of skin is greater in the savanna-dwelling than in the forest dwelling species of frogs, in savanna species there is a predominance of mucous glands, these produce mucous secretions that help the animal in its adaptation to high temperatures and low relative humidity environments (Le Quang Trong, 1975). Mucous production depends directly on gland density and it has been shown that the mucous protects against desiccation. These Mucous glands are important in thermal and water economy relationship of the frog and its environment, mucous discharges aid in the control of body temperature and also maintain the amphibian skin moist for cutaneous respiration.

The secretions produced by serous cutaneous glands in the order Anura exhibit highly variable ultrastructural features (Delfino et al. 1992). Serous storage bodies represent a hetorogenous class of structures ranging from vesicles containing translucent products to dense membrane bounded aggregates; their morphological variation also includes accumulations resembling multivesicular bodies. This heterogenity reflects specific biosynthetic pathways during the post golgian maturation phase which can be easily seen in the premetamorphic stages of development, mature serous products are consistent for each species in each genus investigated (Delfino 1991).

Daly et al. (1987) comment that the wide variability in both composition and function of serous secretions of anuran skin reflects evolution of the survival strategies in the living families. Species from genus Phyllomedusa are known for exhibiting great polymorphism in their sereous gland morphology, studies performed by G. Delphino et al. (1998), show that variation of size and histochemical properties vary among species present in Argentina. Phyllomedusa species possess at least three serous gland types that have been classified on base of morphological and histochemical characterization. Skin permeability has been found to be greatly influenced by cutaneous lipids (Schim and Bardem, 1965). Blaylock et al. (1976) described peculiar glands in Phyllomedusa. These glands named as lipid secreting glands, were proved to be related with regulating evaporative water loss through the skin, frogs from this genus spreads lipids over the body surface using all limbs with a stereotyped whipping behavior (Blaylock et al.1976).

Arboreal hylids are potentially more exposed to dehydrating conditions, thus some authors as Yorio and Bentley (1977), have described much of the adaptations favoring water conservation by the body. Lipid quantity in the ventral skin of Agalychnis dacnicolor is less than in the ventral skin of other anuran species such as Bufo marinus, Rana pipiens and Xenopus laevis. In some Phyllomedusa species (ex. P. bicolor), ossified structures appear as bony spines which project outwards, covered by epidermis and originate from basal osseous plates in the dermis that possess low vascularized regions aiding in water conservation.

The ventral pelvic or inguino-femoral region in anurans has a powerful capacity for water absorption. Habitat and hydration capacity in an anuran can be related to the vascularization of the integument in the pelvic region (Roth, 1973). The skin of the pelvic region is morphologically different from that of other parts of the body, being thinner and well vascularized. The degree of terrestriality of a species seems to be related with the greater intensity of cutaneous vascularization in the pelvic region, this morphological aspect is linked with behavioral postures for rehydration, adaptations that favor positive water flow into the body. Structurally, water absorption pads are configured by small verrucae hydrophilica, a cutaneous structure provided with specific vascular plexa (Drewes et al., 1977). Each verruca is usually composed by a central granular gland, surrounded by four to six mucous glands; Capillary blood vessels of various sizes are distributed over the surface of the verruca, some of these are placed at the base of the sulci, near the epidermis, these sulci store water thus preventing evaporation. Kolbelt and Lisenmair (1986) have described that it is more probable that water absorption is taken place along the sulci than on the surface of the verruca. In the amphibians the presence of capillaries in a sub epidermal position is considered as a primitive character, epidermal capillaries are an adaptation of some terrestrial amphibians to rapid abortion of water (Czopek and Szarsk, 1989 in Toledo et al., 1993).In this thesis morphological aspects of the amphibian skin are discussed based on histological observations performed on light microscopy, a preliminary characterization of dry forest species and comparisons between wet forest populations are shown as a qualitative approach for trait variation.

Figure 1. Strategies often employed by the dry forest community to avoid water stress: A) posture employed by Hyloscirtus sp. An arboreal hylid during mid-day; B) underground retreats may be used to aestivate or as a humid refuge by humboldti; C) Tree trunk cavities are used by arboreal hylids.

Part 2. Body size.

The effect of climate on body size proportions has been studied along aridity gradients and a trend between rainfall, limb length and cranial width has been observed (Lee, 1993). Environmental heterogeneity and body size has been studied in different latitudes under distinct level of analysis. Olalla et al. (2009) performed a community assemblage approach for the variation observed in Brazilian Cerrado anurans, and concluded that water deficit is the only explanatory variable for the observed pattern which dictates that larger body sizes are associated with dryer areas. On the contrary Greene et al (2013), based on 23 years of measurements and skeletochronology on a temperate species found that body size is more related to abundance than to abiotic factors such as rainfall. So patterns have been discussed as being more related with phenotypic plasticity than to a real evolutionary response. As part of this thesis body size was tested among different dry forest related species and their wet forest sister population. Iterspecific analysis was performed for two wet forest-dry forest sister species to test for any phylogenetic trend in body size.

Methods:

Part 1. Histology.

96 skin samples from a total of 27 individuals from 10 species in four families were collected from the Inguinal, ventral and dorsal regions of the frog´s body. Samples were fixed in 10% formalin, dehydrated in ascending series of ethyl alcohols and embedded in paraffin. Transverse skin sections of 7 micrometers were hydrated and stained with Ehrlich´s hematoxylin and Eosine method (1886), this process was carried out by an ICA institute histopathologist. Analysis was performed using optical microscopy and measures were obtained using an ocular micrometer. The work was documented with photographs taken using a digital camera. The examined material belongs to collections performed by the author in the departments of Guajira, Cesar, Atlántico, Cordoba, Bolivar, Cesar, Antioquia and Huila. Measurements are presented as descriptive on base of literature records for cutaneous adaptive structures following Toledo et al. (1993), Mangione et al. (2009); Delfino et al. (1998); Elias et al. (1957) and Perez et al. (1996), Duellman et al. (1986). Nomenclature and morphometric methods follow these authors as well.

Figure3. Species analyzed through histology: Left: Arboreal species, family Hylidae A) Trachycephalus typhonius, B) Hypsiboas crepitans; C) Phyllomedusa venusta. Center: terrestrial Leptodactylidae A) Leptodactylus fuscus; B) Leptodactylus fragilis; C) Leptodactylus bolivianus. Right: terrestrial miscellanea A) Ceratophrys calcarata; B) Rhinella humboldti; C) Pseudopaludicola pusila.

Part 2. Body size.

A total of 427 museum specimens were measured belonging to ten species in four families. The present study is limited to sexually mature males that were distinguished from females by secondary sexual characters like nuptial pads, vocal sacs, and spines or by sexing the individual directly by means of gonad inspection. Only males were selected because of the certainty of discarding juvenile frogs that can´t be easily distinguished from females in many cases. Three measures were taken from each individual: SVL (Snout vent Length), TL (Tibia length) and cranium width (CW), which represents measurements relative to bone structures, whose dimensions do not change dramatically after fixation and preservation. All Measurements were performed with dial caliper accurate to 0.1 mm. A quantitative intraspecific test was performed using the GIS data available for every measured individual in the museum, this made possible the inclusion of an additional analysis with BIOCLIM variables (Bio1, Bio 12 and BIO 15) vs. the distribution of body proportions (TL/SVL; CW/SVL) found in individuals collected from different localities, that were deposited in the amphibian collection at ICN, Instituto de Ciencias Universidad Nacional de Colombia. Interspecific analyses were performed among sister species such as Dendrobates truncatus and D. auratus (Dendrobatidae) and from Trachycephalus typhonius and T. resinifictrix (Hylidae), the remaining species were analyzed independently by intraspecific comparisons between sisters populations.

Figure 2. Measures used to test body size relations

Results

Seven morphological adaptations related with water economy were described in the dermal integument: presence of E-K Layer and calcified layers, Presence of specialized lipid glands; Elevated mucous gland density; epidermal sculpturing and epidermal grooves; Iridiophores, interdependency with the lymphatic system. Specialized vascular plexa and verruca hydrophilica, are differentially distributed in the ventral region within a single species, following apparent geographical patterns. Most of the structures have an asymmetrical distribution along the anuran skin conferring differential properties to ventral, inguinal and dorsal portions of the animal. Terrestrial and arboreal species differ greatly in tegumentary structures and thus were analyzed separately; a descriptive analysis is presented for the three regions explored; Dorsal, Ventral and Inguinal portions of the frog´s skin. Results here presented are from nine species in four families including: Leptodactylidae (Leptodactylus bolivianus, L. fuscus, L. fragilis Pseudopaludicola pusila); Hylidae (Hypsiboas crepitans, Phyllomedusa venusta, Trachycephalus typhonius); Bufonidae (Rhinella humboldti), and Ceratophrydae (Ceratophrys calcarata) a table is presented with the details of the corresponding measurements for each species and population (Table 1).

Dorsal skin:

Leptodactylids possess a simple dorsal epithelium rich in dermal and subdermal mucous glands that resemble mucous gland described for other amphibians (Duellman et al. 1986), these last glands present in both species possess elongated secretory ducts that differ from previously described structures in other Leptodactylids (Figure 1A), other structures that can be observed in Leptodactylids are the presence of pores that interrupt the calcified layer and extend to the surface of the epidermis (Figure 1B). Serous glands are poorly represented in species like Leptodactylus fuscus and L. fragilis, but can be abundant and large in Leptodactylus bolivianus (Table 1.). Mucous glands can be also found related with dorsal folds in L. fuscus (Fig. 5A) and show an aggregate distribution (220 glands in one millimeter of skin). Intersticial spaces product of cell apoptosis can be observed beneath the stratum compactum in L. bolivianus (Fig. 4A) Leptodactylids possess a clearly defined calcified layer (CL) that can be continuous as in L. fragilis (Fig 5B) or poorly interrupted as in L. bolivianus (Fig. 4A). Dorsal skin from recently emerged post metamorphic individuals (Dorsal #55 in Table 1) show the early appearance of the calcified layer in the dorsal skin of this species. Dry area Leptodactylids have a thicker skin than mesic forms but a considerably thinner stratum corneum (Table 1.)

A B

Figure 4. A) Leptodactylus bolivianus, dorsal skin showing the presence of large serous glands, a thick calcified layer and sub dermal mucous glands with elongated secretory ducts; B) Leptodactylus fragilis, dorsal skin showing pore that interrupts the calcified layer. Bar = 100 µm.

A B

Figure 5. A) Leptodactylus fuscus, dorsal fold sowing high density of mucous glands; B) Leptodactylus fragilis, dorsal skin showing the presence of mucous glands and poorly interrupted calcified layer. Bar = 100 µm.

Hylids show a greater diversity of dorsal glands than other dry forest frogs, skin from Phyllomedusa venusta has three types of serous glands and one type of mucous glands. Serous glands are specialized syncythia that differ greatly in morphology and secretory products which stain differently when running a routine dye. Two types of serous glands are related to the production of proteinous substances employed in defense against predators and infections, these are the glands defined as Ia and Ib as seen in figure 6 A. Histological nomenclature used for Phyllomedusa was based on Delfino et al. (1998). Histologically type I glands produce large collection of sphaeroidal densely stained granules or translucentlls vesicles, this glad type has a great morphological variation and is differentiated from other gland types by having a secretory compartment ensheathed by a contractile layer of mioepithelial cells, but for its upper pole, several layers of undifferentiated cells rest in the Syncythium. These are both adenoblasts and myoblasts involved in the cycling of the secretory unit and myoepithelial layer. The neck is localized at the boundary between the epidermis and dermis and it is encircled by chromatic units, it holds a slender cavity joining the exiguous lumen of the secretory unit to the gland duct, which is entirely contained within the thickness of the epidermis. Specialized glands related to the production of lipid secretions are defined in the literature as Type II glands, these manufacture discrete secretory bodies appear weakly stained, type II glands possess a relatively larger lumen than type I glands. The lumen of these glands varies in width and is bounded by a thin layer of flat cells with nuclei located at the apex of the secretory unit; cells with a high nucleo-plasmatic ratio are stratified form the gland neck. This species additionally possess a great number of Iridiophores that form a layer next to chromatophores just beneath the epidermis. These are deposit centers of urate salts from the nitrogen metabolism (Fig. 6 A). Particularly this species lacks a calcified layer, but other hylids such as Trachycephalus typhonius and Hypsiboas crepitans have the presence of a poorly interrupted layer, the first species has a greater degree of calcification of the layer reaching width of up to 24.7 µm (Table 1). T. typhonius shows a great abundance of serous glands that produce proteinous secretions (Fig 7.)

A B

Figure 6. A) Dorsal section from Phyllomedusa venusta showing gland diversity; serous glands Ia and Ib, and mucous glands (M), and iridiophores (I). B) Dorsal section from P. venusta showing a detail of a Type II lipid gland that has a larger lumen and structured secretory vesicles can be observed. Bar=100 µm.

Figure 7. Dorsal skin from Trachycephalus typhonius, showing the presence of a poorly interrupted calcified layer (CL) and the presence of large serous proteinous substance secreting glands (SG). Bar = 100 µm.

Other Terrestrial anurans such as bufonid Rhinella humboldti possess a cornified epidermis with thickened regions that form worts and spines (Fig. 8B); bufonids show epidermal sculpturing and regions with low gland density, including mucous glands that are poor and spaced, thus bufonid skin is very dry. Serous glands produce proteinous secretions that are more concentrated over the extension of the parotid gland. Calcium layer was only found in the dorsal skin of humid area individuals and was heavy calcified in some regions of the dorsum, reaching 24.7 µm in thickness (Figure 8A). Arid area individuals lacked the calcified layer in every region of the body.

Terrestrial burrower, Ceratophrys calacarata possess a unique character in its dorsal skin, the presence of a thick calcified structure called E-K layer, structure associated with adjacent living cells (Fig 8C), and this tissue may be 7.49-49.4 µm thick. Skin in this species is also very thick (128-350 µm). The E-K layer is poorly interrupted and is only present in the dorsal skin.

A B C

Figure 8. A) Dorsal Skin from Rhinella humboldti showing calcified layer (CL) and the presence of large serous glands (SG); B) Rhinella humboldti dorsal skin showing a spine formed in the outer epidermis by keratinocytes. C) Dorsal skin from Ceratophrys calcarata, showing the presence of an E-K layer and mucous gland (M). Bar = 100 µm.

Ventral skin:

Different structures may be found in amphibian ventral skin, some species like Leptodactylus bolivianus show the presence of a poorly structured ventral region with large blood vessels (L) Fig 9B. The ventral lymph system lies just beneath the dermis. Phyllomedusa venusta water absorption pads are located only in the ventral skin, verruca hydrophilica are present along the venter and Type II lipid glands are completely absent (Fig 9A). Species such as Hypsiboas crepitans have a specialized ventral water absorption pad and it is configured by numerous pronounced and vascularized verrucae (Fig 9C.). Species such as Ceratophrys calacarata lack absorption structures in the ventral skin. Ventral skin in Pseudopaludicola pusilla differs little from the dorsal skin and has no specialization to facilitate water absorption. Rhinella humboldti possess well defined and vascularized ventral verruca.

A B C

Figure9. A) Ventral skin from Phyllomedusa venusta showing the presence of water absorption verrucae, B) ventral skin from Leptodactylus bolivianus showing a well extended vascular plexa (L), related to the ventral lymphatic sac; C) ventral skin from Hypsiboas crepitans showing the presence of well vascularized sulci (S).

Inguinal skin:

Inguinal skin shows fairly specialized water absorbtion pads in most of the species. The ventral region of the thigh In Leptodactylus fuscus (Fig. 10A.),and most of leptodactylids show the distribution of vascular plexa formed by epidermal capillaries (EC), that configure pronounced and heavy vascularized verruca, epidermal capillaries are located in the base of the verruca and large mucous glands occupy most of the stratum spongiosum (Fig 10A). Verrucae from arboreal species such as Trachycephalus typhonius differ by having all epidermal cappilaries in the distal lining of the sulcus (Fig. 10B) and by having a greater degree of vascularization on the verruca as shown for Trachycephalus resinifictrix in figure 10 D. Terrestrial, Ceratophrys calacarata possess a complex vascular plexa, distributed along the whole sulcus, some mucous glands (Fig. 10A) are also present. This type of verruca is the only type of water absorption structure found in the complete ventral region of this frog. Arboreal Hypsiboas crepitans possess small verruca hydrophilica along the underside of the thigh, but are not as pronounced and specialized as the ventral structures; Phyllomedusa venusta, another arboreal species, lacks inguinal verruca and has an important number of type II lipid glands in this section of the skin, indicating that this is not a water absorption zone. The bufonid, Rhinella humboldti also show the presence of inguino-femoral verruca.

A B

C D

Figure 10. A) Verrucae hydrophilica in Leptodactylus fragilis, showing the presence of vascular plexa (EC) and large mucous glands; B) Verruca hydrophilica in Trachycephalus typhonius, showing the presence of vascular plexa in the distal portion of the sulcus (EC); C) Verruca in Ceratophrys calacarata showing an organized structure for water absorption, configured by ascending epidermal capillaries along the inner portion of the sulcus (EC) and the presence of few mucous glands (M). D) Verruca in arboreal Trachycephalus resinifictrix showing important vascular plexa distributed along the verruca (EC). Bar = 100 µm. Table 1. Measurements of integument and structures in seven species of seasonally dry forest frogs and comparisons among sister populations. Leptodactylus bolivianus : Leptodactylidae Location of TOTAL Continous Mucuous Granular Mucous Granular verruca skin Region Thickness Stratum Stratum Stratum Calcified or gland gland gland gland hydrophilic Locality and code µm corneum Epidermis spongiosum compactum layer interrupted density density dimensions dimensions a Dorsal.#1 333.45 0.247 19.76 24.7 106.21 17.29 continous 18/1000 3/1000 93.1X196 34.3X132.3 Cesar, Dry Ventral.#2 419.8 1235 24.7 0 296.4 0 0 0 0 0 0 Forest Inguinal. #3 209.95 3705 29.64 49.4 86.45 4.94 Interrupted 13/1000 0 61.75X34.58 0 Inguinal Dorsal.#37 218.85 4.94 49.4 86.45 79.04 29.64 Continous 53/9,015.5 68/9,015.5 49.4X49.4 74.1X98.8 Huila, Ventral.#38 326.04 7.41 49.4 128.44 148.2 25 Interrupted 75/5,693.35 2/10000 56.81X46.93 86.45X83.98 Garzon, Dry forest Inguinal.#39 296.4 2.47 24.7 25 244 0 0 35/1000 0 37.05X49.4 0 Inguinal Dorsal# 55 121.14 0.98 11.76 34.3 74.1 3.93 Continous 74/1000 0 24.7x24.7 0 Cesar, Dry Ventral#56 148.2 9.88 37.05 37.05 51.87 0 0 85.5/1000 0 49.4X37.05 0 Forest Inguinal#57 83.98 2.47 24.7 24.7 32.11 0 0 42/1000 0 19.76x17.29 0 Inguinal Leptodactylus fuscus : Leptodactylidae Arauca Dorsal.#48 153 3.7 24.7 49.4 106.21 2.47 Interrupted 54.34 /

Ventral. #47 116.09 4.94 29.64 37.05 44.46 0 0 64/1000 0 /

Inguinal# 46 160.55 2.47 14.82 49.4 98.8 0 0 / Inguinal Dorsal. #49 106.21 24.7 61.75 0 0 40/1000 1/1000 / Ventral #50 125.97 2.47 29.64 61.75 32.11 4.94 Interrupted 83/1000 1/1000 54.34x61.75 / Huila, Inguinal # Garzon 51 / / Inguinal Dorsal. #94 271.7 7.41 34.58 148.2 86.45 7.41 Continous 220/1000 / /

Ventral. #95 210.76 0.81 49.4 61.75 98.8 2.47 Interrupted / 49.4x49.4 /

Cesar Inguinal #96 106.21 2.47 29.64 69.16 49.4 0 0 / / Inguinal Dorsal. #97 169.84 7.41 29.64 71.63 69.16 9.88 Continous 73/1000 / 61.75x74.1 / Ventral.#98 123.95 7.41 37.05 24.7 54.3 12.35 Interrupted / 49.4x24.7 / Inguinal. Antioquia #99 113.62 2.47 44.46 24.7 66.69 4.94 Interrupted / / Inguinal Trachycephalus typhonius : Hylidae Dorsal.#7 419.9 7.41 37.05 148.2 217.7 24.7 continous 49/1,000 17/1,000 86.45X98.8 358.15X382.85 0 Tubará/ Dry Ventral.#8 359.2 2.47 37.05 358.15 227.24 12.35 Interrupted 21/1,000 7/1,000 74X37.05 358X148.2 Ventral forest Inguinal. #9 560.69 2.47 61.75 135.85 407.55 9.88 Interrupted 16/1,000 6/1,000 81.51X61.75 234.65X111.15Inguinal Phyllomedusa venusta : Hylidae Dorsal. #19 259 7 37 111 136 / / / / / /

Ventral. #20 252 10 37 86 124 / / / / / / Antioquia, Inguinal. Dry forest #21 254 7 49 136 62 / / / / / / Ventral Dorsal . #25 345.8 4.94 24.7 160.65 135.85 0 0 30/1000 10/1000 79.09x61.75 296.4x239.59

Ventral. #26 358.15 2.47 37.05 197.6 123.5 0 0 47/1000 74.1x209.95 135.85x185.25 Cesar, Dry Inguinal. Forest #27 247 4.94 24.70 37.05 49.4 0 0 15/1000 6/1000 / / ventral Pseudopaludicola pusilla : Leptodactylidae / 0 Dorsal. #40 135.85 2.47 12.35 41.99 86.45 0 0 122/1000 12/1000 37.05x24.7 / 0 Cesar, Dry forest Ventral. #41 88.92 2.47 24.7 24.7 37.05 0 0 84/1000 2/1000 / / 0 Inguinal. #42 54.34 4.94 13/1000 / / / 0 Hypsiboas crepitans: Hylidae Dorsal. # 13 350.74 4.94 49.4 140.79 155.61 7.41 Interrupted 21/1000 18/1000 106.21x123.5136x148 0

Ventral. #14 422.37 2.47 74.1 247 98.8 0 0 22/1000 1/1000 / / Ventral Cesar, Dry Inguinal. forest #15 340.86 7.41 111.15 123.5 98.8 0 0 15/1000 2/1000 / / Inguinal Dorsal.#58 229.71 4.94 37.05 98.8 88.92 4.94 Interrupted 44/1000 24/1000 / / 0

Cundinamar Ventral. #59 380.6 4.94 79.04 74.1 0 0 32/1000 8/1000 61.75x61.75 83.98x98.8 Ventral ca, Wet Inguinal. Forest #60 2.47 17.29 81.51 49.4 0 0 31/1000 21/1000 49.4x61.75 123.5x61.76 Inguinal Rhinella humboldti : Bufonidae 0 / / / / Dorsal # 43 412 12 49 178 173 0 / / / / Huila, Dry Ventral #44 191 5 25 25 136 0 / / / / Ventro- forest Inguinal # 45 210 2 25 49 148 0 / / / / Inguinal Dorsal. # 64 340.86 14.82 37.05 135.85 153.18 24.7 0 / / / / Ventral. # 65 271.7 2.47 24.7 123.5 123.5 0 0 / / / / Antioquia, Inguinal. Ventro- Wet forest #66 338.39 2.47 29.64 148.2 160.55 0 0 / / / / Inguinal

Part 2. Body size:

Results for the test of linearity show a direct response for two of the three Bioclim variables tested Bio 15 and Bio12, (Precipitation seasonality and Mean annual precipitation), Bio 1 (Maximal temperature) showed no clear trend between the species analyzed. Only two species showed a linear response: The arboreal hylid Trachycephalus typhonius and the diurnal dendrobatid Dendrobates truncatus. T. typhonius shows a positive linear relation between cranium width proportions and precipitation seasonality (Figure 11, Table 2). D. truncatus shows an inverse linear relation between increased precipitation seasonality and tibial length proportions and a positive relation between increased annual precipitation and tibial length proportions (Fig12, Table 2). None of the remaining species showed a significant linear relationship between body proportions and precipitation variables.

Table 2. P value for each test of linearity performed between bioclim variables, Bio 15 and Bio 12 and considered body proportion (Tibial length and Cranium width)

BIO 12/ TL Bio 12 /Cr Bio 15/ LT Bio 15/Cr Species P value P value P value P value Hypsiboas_pugnax 0.589 0.187 0.7878 0.134 Leptodactylus_fuscu s 0.898 0.189 0.208 0.222 Leptodactylus_insula rum 0.293 0.785 0.565 0.357 T. typhonius 0.2957 0.13319 0.711 0.027 T. typhonius and T. resinifictrix 0.129039 0.07097 0.687552 0.47 Dendrobates_trunca tus 0.05589 0.683 0.4702 0.523 D. truncatus and D. Auratus 0.0488 0.689 0.0394 0.519 Pleurodema_brachyo ps 0.129039 0.07097 0.2138 0.729

Figure 11. linearity test for bioclim variable BIO 15 and Cranium width proportion (Cr).

Figure 12. Linear regression for annual precipitation and seasonality vs Tibial length proportions (TL) in the diurnal species Dendrobates truncatus.

Discussion

Arboreal and terrestrial species differ in dermal tegument configuration and structure; terrestrial species show specialized ventral and inguinal patches to actively absorb water from the substrate, specialized vascular plexa are differently distributed in each species, showing anatomical differences in the structure of verruca Hydrophilica. In terrestrial species like frogs from genus Leptodactylus, ventral skin may not have specialized rehydration structures, this was also evidenced for Pseudopaludicola and Ceratophrys, in these genera verrucas are restricted only to the inguinal portion of the animals. Some leptodactylids such as L. bolivianus show a simplified ventral skin associated with the ventral lymph sac, this interaction between systems may be related to water conservation, as the lymphatic system serves as a reservoir of water for the blood in case of dehydration (Hillman et al. 2005). Other terrestrial species as bufonids have drinking water pads in both ventral and inguinal position. Terrestrial species may possess a thicker and more keratinized epidermis than arboreal species, especially those burrowing species. Species like Rhinella humboldti show thicker ventral and dorsal skin in wet forest populations but similarly thinner ventral and dorsal skin in individuals from the dry forest. Navas (2004), has proposed that this condition enhances facultative water absorption, even from the dorsal portion, while the animal is underground (Fig 1B). Burrowing species as Ceratophrys calcarata possess a wide calcified layer (E-K) that has been argued to be a physical barrier that reduces water evaporation, but it has also been defined as a plesiomorphic character within the genus, with no clear function in water conservation (Mangione et al. 2011), Calcified layers are continuous in Leptodactylus species, with little interruptions except for gland ducts and pores (Fig. 4A). The calcified layer appears early in the development of L. bolivianus, being present in recently emerged metamorphs, the calcified layer only appears to be continuous and thick in the dorsal portion off all species that present it (Table 1). Mucous gland density can be greater in terrestrial frogs from dry zones and has a function in both keeping the skin moist for cutaneous gas exchange and in the case of glandular tissue found in dorsal folds, as a predator deterrent. Mucous glands are also associated with verruca hydrophilica in most of the species; these glands carry an important function as secretory units that prevent water loss when the frog is placed in dehydrating substrates.

Arboreal species possess a greater diversity of serous glands and some species have evolved specialized lipid secreting units (Type II glands). Phyllomedusa venusta shows an increased abundance of type II glands in the dorsal skin that diminishes when approximating the ventral skin, the presence of type II glands in the inguinal skin and the presence of smaller poorly vascularized verruca show that the inguinal portion of the animal is not specialized in water absorption, suggesting that the venter fulfills rehydration needs. Phyllomedusa venusta shows large and abundant iridiophores in the dorsal skin, structures that store urate salts and are thus related with the physiology of water economy. Hylids Hypsiboas pugnax and species of Trachycephalus show the presence of both ventral and inguinal verruca, but both differ structurally; specialized organization of the vascular plexa can be seen in inguinal verruca. Calcified layers appear in some arboreal hylids such as Trachycephalus and Hypsiboans but are very interrupted, Phyllomedusa venusta, lacks a calcified layer.

Many features of the dermal organization vary geographically between populations and may account for variation induced by environmental conditions imposed through ontogeny, description of this variation still remains preliminary, and thus no sufficient evidence exists to support an adaptationist hypothesis that accounts for the observed variation. The analyzed species show a close relation between dermal organization and their natural history; following three discrete eco types: Arboreal species, fossorial and terrestrial species with some degree of variation within each ecotype. Tests performed over body size proportions and environmental variables show no evidence to support Bergmann´s rule, body proportions are not necessarily related with environmental variables such as temperature, Precipitation or seasonality, thus few species and population follow a clear trend. Only two species showed a significant linear relation influenced by precipitation or seasonality in rainfall. The one that best shows a pattern is Dendrobates truncatus a diurnal species, who principally inhabits stream forest. Populations from this species inhabiting places with very low values of mean annual precipitation, have smaller tibia proportions compared to those populations in places with high rainfall (Fig 12A). A similar pattern occurs when testing against precipitation seasonality; were tibial lengths are shorter with increased values of variation (Fig 12 B). Cranium width proportions showed no relationship for this species, suggesting that the width of the animal´s body is not affected by the patterns in rainfall. Diurnality can be an important aspect of this frog´s life history, and it could explain intraspecific variation in body size, as it implies exposure to higher temperatures and greater water loss as evotranspiration is greater, when compared to the conditions experimented by species with enhanced nocturnality (Navas, 2004). A second species that showed a positive relationship between body proportions and precipitation variables is Trachycephalus typhonius, a positive relation between the coefficient of variation on the precipitation seasonality and cranium width proportions exists, showing greater proportions of the skull related to extremely temporal sites with prolonged dry seasons. A possible explanation for this last trend is difficult to interpret because of the evident lack of co-osification found on this specie´s head. Further research must consider that though much of the variability is generated by phenotypic response to environmental extremes a possible evolutionary trend may exist when comparing sister species, much more information on this direction is needed.

Acknowledgements: Very grateful with Andrew J. Crawford for his patience and for his direction in the development of this thesis, this project began from scrap, and evolved slowly thanks to conversations with him during many years; also with my potential co director Santiago Madriñan who encouraged part of this work. Special thanks to Dr. Emilio Realpe, for his help, advice and observations; professor Jhon D. Lynch at ICN for letting me measure frogs and for his very constructive advice that saved many aspects of this work; Carlos Navas also has inspired this work and his comments have been important pillars of this research and special thanks to Fercho at ICA for his wonderful work on the slides. Fish vet Miguel Mendoza was very helpful with his ideas and advice during the exploratory phase of the histological work and to my folks Hugo and Luz Mery for helping me in every step of the way

References:

Barbeau, T.R. and H.B. Lillywhite. 2005. Body wiping behavior sassociated with cutaneous lipids in hylid tree frogs of Florida. Journal of Experimental Biology, 208, 2147– 2156.

Bentley, P.J. 1966. Adaptations of amphibia to the arid enviroments. Science, 152: 619-623.

Blaylock, L.A. Ruibal, R. and K. Platt-Aloka. 1976. Skin structure and wiping behavior of Phyllomedusine frogs. Copeia, 1976, 283-295.

Canziani, G. A. and M.A. Cannata .1980. Water balance in Ceratophrys ornata from two different environments. Comparative Biochemical Physiology, 66A, 599-603.

Christensen, C.U. 1974. Adaptations I water economy of some anuran amphibian. Comparative Biochemical Physiology, 47 A, 1035-1049. Cuentas, D. et al. (2002). Anuros del departamento del Atlántico y Norte de Bolívar. Publicación de la Universidad del Atlántico, Barranquilla, Colombia. Delfino, G. Alvarez, B.B. Brizzi, R. and J.A. Cespedez. 1998. Serous cutaneous glands of argentine Phyllomedusa Wagler 1830 (Anura Hylidae): Secretory polymorphism and adaptive plasticity. Tropical zoology, 11, 333-351.

Drewes, R.C. Hillman, S.S. Puntam, R. W. and O.M. Sokol. 1977. Water, nitrogen and ion balance in the African tree frog Chiromantis petersi Boulenger (Anura; Rhacophoridae) with comments on the structure of the integument. Journal of Comparative Physiology, 116, 257-267.

Duellman, W. E. and L. Trueb. 1986. Biology of Amphibians. McGraw-Hill Book Company, New York. Elias, H. and Shapiro, J. 1957.

Duellman, W. 1978. Tropical herpetofaunal communities: Patterns of community structure in neotropical rainforests. In Vertebrates in complex tropical systems. Edited by Mireille L Harmelin-vivien, Francoise bourliére, France. Elkan, E. 1968. Mucopolysacarids in the anuran defense against desiccation. Journal of Zoology London, 155, 19-53.

Faivovich, J. Haddad, C. F. B. Garcia, P. C. A. Frost, D. R. Campbell J. A. and W. C. Wheeler. 2005. Systematic review of the frog family Hylidae, with special reference to Hylinae: a phylogenetic analysis and taxonomic revision. Bulletin of the American Museum of Natural History, 294,1–240. Galvan-Guevara, S., et al. (2009). Herpetofauna registrada para el área de influencia de la reserva forestal protectora serranía de Coraza, Coloso, Sucre, Colomba. Revista Colombiana de Ciencia Animal, 1 (2). Greene, D.M. and J. Middleton. 2013. Body size varies with abundance not climate in an amphibian population. Ecography, 36, 947-955.

Hedges,S.B. and M.P. Heinicke, Molecular phylogeny and biogeography of West Indian frogs of the genus Leptodactylus (Anura: Leptodactylidae). Molecular Phylogenetics and Evolution, 44, 308–314. Hillman, S.S. Withers, P.C. Hedrick, M. S and R.C. Drewes. 2005. Functional Roles for the Compartmentalization of the Subcutaneous Lymphatic Sacs in Anuran Amphibians. Physiological and Biochemical Zoology, 78(4), 515-523. Infante-Rivero, E. 2009. Anfibios y reptiles de la guajira venezolana. Boletin del centro de investigaciones biológicas. Universidad de Zulia, Maracaibo, Venezuela, 43(2), 263-277. Kobelt, F.L.K. 1992. Adaptations of the reed frog hyperolius viridiflavus (amphibia, anura, hyperoliidae) to its arid environment: The iridiophores in the skin as radiation reflectors. Journal of Comparative Physiology, 162B, 314-326.

Kobelt, F. and K.E. Linsenmair. 1986. Adaptations of the reed frog Hyperolius viridiflavus (Amphibia, Anura, Hyperolidae) to its arid environment. I. The skin of Hyperolius viridiflavus nitidulus in wet and dry season conditions. Oecologia, 68, 533-541

Le Quang Trong, Y. 1971. Etude de la peau et des glandes cutanées de quelques amphibiens du genre Phrynobatrachus. Bulletin of the french institute for northern African natural sciences, 33, 987-1025.

Le Quang Trong, Y. 1975. Etude de la peau et des glandes cutanées de quelques amphibiens du genre Ptychadema. Ann. Univ. Abijan, Ser. Ecol, 8, 31-52.

Lillywhite, H. B. 2006. Water relations of tetrapod integument, Review. Journal of Experimental Zoology, 209, 202–226. Lynch, J.D., 2006. The tadpoles of frogs and found in the lowlands of northern Colombia. Revista de la Academia Colombiana de Ciencias, 30(116). Lynch, J. D., et al. 1997.Biogeographic patterns of Colombian frogs and toads. Revista de la Academia Colombiana de Ciencias, 21(80). Mangione, S. Garcia, G and O. Cardozo. 2011. The Eberth-katschenko layer in three species of cetatophryines anurans (Anura: Ceratophrydae). Acta Zoologica, 92, 21-26.

Yorio, T and P.H. Bentley. 1977. Asymetrrical permeability of the integument of tree frogs (Hylidae). Journal of Experimental Biology, 67, 197-204.

McClanahan, L. Ruibal, R. and V. H. Shoemaker. 1994. Frogs and toads in deserts. Scientific American. 270 (3), 82-88.

Navas, C.A. 1996. Metabolic physiology, locomotor performance, and thermal niche breadth in neotropical anurans. Physiological Zoology, 69, 1481-1501.

Navas, C. A. et al. 2004. A preliminary assessment of anuran physiological and morphological adaptation to the Caatinga, a Brazilian semi-arid environment. International Congress Series, 1275, 298 – 305. Olalla-Tárrraga, M.Á. Diniz-Filho, A.F. Bastos, R.P. and M.Á. Rodriguez. 2009. Geographical body size gradients in tropical regions: Water deficit and anuran body size in the Brazilian Cerrado. Ecography, 32, 581-590.

Ruibal, R. and V.H. Shoemaker. 1984. Osteoderms in anurans. Journal of Herpetology, 18, 313-328. Supplementary information:

Significance (p values) obtained in testing of linearity between body proportions versus 3 different bioclim variables: maximal temperature, annual precipitation and seasonality in precipitation.

Temp Máxima

Temperatura LT Temperatura Craneo Valor P Valor P Hypsiboas_pugnax 0.95033 0.19386 Leptodactylus_fuscus 0.7018 0.9424 Leptodactylus_insularum 0.6777 0.6892 T. typhonius and T. resinifictrix 0.5639 0.1658 T. typhonius 0.88 0.2648 Dendrobates_truncatus 0.98 0.64832 D. truncatus and D. Auratus 0.9151 0.7652 Pleurodema_brachyops 0.7219 0.7399

Precipitación Anual

Precipitación LT Precipitación Craneo Valor P Valor P Hypsiboas_pugnax 0.589 0.187 Leptodactylus_fuscus 0.898 0.189 Leptodactylus_insularum 0.293 0.785 T. typhonius 0.2957 0.13319 T. typhonius and T. resinifictrix0.129039 0.07097 Dendrobates_truncatus 0.05589 0.683 D. truncatus and D. Auratus 0.0488 0.689 Pleurodema_brachyops 0.129039 0.07097

Precipitación Estacionalidad

Estacionalidad LTEstacionalidad Craneo Valor P Valor P Hypsiboas_pugnax 0.7878 0.134 Leptodactylus_fuscus 0.208 0.222 Leptodactylus_insularum 0.565 0.357 T. typhonius and T. resinifictrix 0.711 0.027 T. typhonius 0.687552 0.47 Dendrobates_truncatus 0.4702 0.523 D. truncatus and D. Auratus 0.0394 0.519 Pleurodema_brachyops 0.2138 0.729